Jurnal RESTI (Rekayasa Sistem dan Teknologi Informasi)
Not a member yet
1071 research outputs found
Sort by
A New Framework for Dynamic Educational Marketing Segmentation in Student Recruitment: Optimizing Fuzzy C-Means with Metaheuristic Techniques
Full text linkAn effective educational marketing strategy requires accurate school segmentation to enhance new student recruitment. Traditional segmentation methods such as K-means are often used, but they have limitations in capturing the flexibility of school characteristics. Fuzzy C-Means (FCM) offers a more adaptive approach by allowing each school to simultaneously have a degree of membership in several clusters. However, the performance of FCM highly depends on determining parameters such as the number of clusters (k) and the level of fuzziness (m), which are not always optimal when determined manually. This study develops a new framework for dynamic educational marketing segmentation in student recruitment by optimizing FCM using three metaheuristic techniques: Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Differential Evolution (DE). Performance was evaluated using the Fuzzy Silhouette Index (FSI). The experimental results showed that DE yielded the best results with the highest FSI value (0.8023), producing eight main clusters based on the Recency, Frequency, and Monetary (RFM) model. Based on the clustering results, a personalized and adaptive marketing strategy was designed to enhance the effectiveness of student recruitment. The proposed framework enhances segmentation accuracy and supports the implementation of dynamic data-driven marketing in the context of higher education. This study also opens new directions for educational data mining research and machine-learning-based marketing strategies
Performance Comparison of Monolithic and Microservices Architectures in Handling High-Volume Transactions
Full text linkMonolithic and microservices are two distinct approaches for designing and developing applications. However, these architectures exhibit contrasting characteristics. In monolithic architecture, all components of an application form a unified entity with closely interconnected parts, whereas microservices decompose an application into independent, lightweight services that can be developed, deployed, and updated separately. Microservices are often regarded as superior to monolithic architectures in terms of their performance. This study aims to compare the performance of monolithic and microservices architectures in handling a high volume of transactions. It is important to observe how the two architectures behave when handling transactions from a large number of concurrent users. A prototype of an online ticketing system was implemented for both architectures to enable comparative analysis. The selected performance metrics were response time and error rate. The experimental results reveal that under high-load conditions, microservices outperform monolithic architectures, demonstrating 36% faster response times and 71% fewer errors. However, under overload conditions—when CPU usage exceeds 90%—the performance of microservices degrades significantly. This does not imply that microservices cannot handle a large number of concurrent users but highlights the necessity for enhanced resource management
Optimizing a Hybrid Deep Learning Model for DDoS Detection Using DBSCAN and PSO
No full textThis study proposes a hybrid deep learning approach that combines Gated Recurrent Units (GRUs) and Convolutional Neural Networks (CNNs) for Distributed Denial of Service (DDoS) cyberattack detection. The model, called DBSCAN–GRU–CNN, uses density-based clustering (DBSCAN) to select relevant features and reduce execution time. The dataset for this study was obtained from live penetration testing, where a series of simulated attacks was performed on a monitored network. To evaluate the performance of the proposed model, several comparison models were used, including DBSCAN–GRU–CNN (Single Hidden Layer), DBSCAN–GRU–CNN (Double Hidden Layers), DBSCAN–GRU–CNN (With Regularization), DBSCAN–GRU–CNN–PSO, GRU–CNN, GRU–CNN (With Hyperparameter Tuning), and Random Forest (Tuned Model). Variations of the model tested were made by adding hidden layers, regularization, optimization with Particle Swarm Optimization (PSO), and hyperparameter tuning. Experimental results show that the DBSCAN–GRU–CNN–PSO model provided optimal performance with a 99.3% accuracy, a 99% precision, a 98.9% recall, and a 99% F1-score, while the model with hyperparameter tuning achieved a 99% accuracy. By adding PSO, the model achieved optimized weights, better generalization, and excellent accuracy in DDoS detection
Analysis of the Impact of Backpropagation Hyperparameter Optimization on Heart Disease Prediction Models
No full textHeart disease is a major global health issue, highlighting the need for early and accurate prediction to reduce complications and improve patient outcomes. The Backpropagation Neural Network (BPNN) is a widely used method for heart disease prediction, but its performance relies heavily on proper hyperparameter selection, including neuron count, activation function, optimizer, and batch size. This study analyzed the impact of hyperparameter optimization on BPNN performance. A standard BPNN model was compared with an optimized version, where key hyperparameters were fine-tuned to enhance predictive accuracy and stability. Both models were trained and tested on the same dataset, and their performance was evaluated using Accuracy, Precision, Recall, Mean Squared Error (MSE), and Mean Absolute Error (MAE). The results show that the optimized model achieves a slightly better accuracy (99.11% vs. 99.09%) and lower error rates (MSE and MAE of 0.0089 vs. 0.0091). It also demonstrates higher precision, reflecting an improved capability in correctly identifying heart disease cases. Although the performance gap was small, the optimized model showed a more balanced and consistent outcome. These findings highlight the importance of hyperparameter tuning for improving neural network models for medical prediction. This study contributes to the development of more accurate and reliable AI tools for the early diagnosis of heart disease. Future studies may apply advanced optimization techniques, such as Bayesian Optimization or Genetic Algorithms, and use larger and more diverse datasets to enhance model generalization
Comparing Data Preprocessing Strategy on T5 Architecture to Classify ICD-10 Diagnosis
No full textManual ICD-10 coding in healthcare systems remains time-consuming, error-prone, and inefficient, particularly in resource-constrained settings. This study investigates the effect of various preprocessing strategies on the performance of the Text-to-Text Transfer Transformer (T5) model for primary diagnosis classification using structured clinical data. A total of 7,263 clinical records were collected from two high-density regions in Bali (Badung and Gianyar) between January 2023 and March 2024, then converted into descriptive text prompts for model training. Four experimental scenarios combined variations of input features and label configurations, comparing T5 with Oversampling against T5 with Easy Data Augmentation (EDA) plus Oversampling. Results showed that T5 with Random Oversampling consistently outperformed the EDA-based configuration across all scenarios, with performance gaps ranging from 8% to 19%. Scenario 4, which excluded body system features and the semantically overlapping E860 label, achieved the highest balance, reaching 84.7% accuracy, 85.1% precision, 84.7% recall, and 84.3% F1-score. Conversely, the EDA-based approach reduced training time by up to 72%, indicating a clear trade-off between performance and efficiency. Both configurations frequently misclassified semantically similar codes within the same ICD-10 categories, underscoring the difficulty of distinguishing clinically related diagnoses. Overall, the results suggest that careful selection of preprocessing strategies can enhance transformer-based medical text classification, while striking a balance between model performance and training efficiency. This work may serve as an initial reference for developing more efficient semi-automated medical coding systems in the Indonesian regional healthcare context
Improving Frame-based Engagement Classification in E-Learning Using EfficientNet and Normalized Loss Weighting
Full text linkEngagement can be defined as how individuals are involved in and interact with a task that requires attention and emotional conditions. Engagement is an affective state positively correlated with learning processes. Engagement along with other affective states, such as boredom, confusion, and frustration must be analyzed to identify students’ learning behavior. Implementing proper prevention by measuring student engagement levels could increase students’ learning intake. Such implementation involves building an effective feedback system or rearranging the learning design. Several researchers have proposed deep-learning approaches using the DAiSEE dataset to classify student engagement levels. In addition, previous studies utilized various loss functions equipped with class weighting to assign higher importance to the minor classes, which are low and very low engagement classes. Most of the state-of-the-art models achieved high accuracy, but the f1-score was still low because of the minor class struggle. This research tries to solve engagement level classification on imbalance conditions by proposing a normalized loss function weighting based on the Inverse Class Frequency formula based on each class’ instances to give more importance and focus to the classes and trained on Vanilla EfficientNet model rather than experimenting on more advanced model to keep the efficient and suit the memory constraint on the e-learning implementation. Based on the conducted experiments, the normalized ICF obtained the highest accuracy of 51.64% and weighted f1-score of 50.86%, which is superior to the standard ICF performance, which received 50.32% accuracy and weighted f1-score of 50.49% using the same settings
Stunting Prediction Modeling in Toddlers Using a Machine Learning Approach and Model Implementation for Mobile Application
Full text linkChildren’s health and development are critical for maintaining national productivity and independence, with stunting being a major concern. Stunting, a form of malnutrition, impairs growth and development, affecting millions of people globally, including a significant number in Indonesia. This study addresses the challenge of stunting by developing a predictive model using machine learning techniques to forecast stunting risks based on public health data. The literature review section discusses the factors that influence stunting, and these factors are used as features to build a stunting prediction model. Then the features were used to build a model with three machine learning algorithms Extreme Gradient Boosting (XGBoost), Random Forest, and K-Nearest Neighbor (KNN) to build and evaluate models that predict stunting. The models were trained and assessed using public datasets and the most effective algorithm was integrated into a mobile application for practical use. The results indicate that the XGBoost model outperforms the other models with an accuracy of 85%, making it the optimal choice for implementation in a mobile application. The next-best model is selected to be implemented through a mobile application so that users can directly use the model that has been built. This application aims to enhance early detection and intervention efforts for stunting, potentially improving child health outcomes and contributing to long-term productivity by building predictive models and implementing the models into a mobile application. This study contributes to the implementation of models built using public data for application in mobile applications
Enhancing Areca Nut Detection and Classification Using Faster R-CNN: Addressing Dataset Limitations with Haar-like Features, Integral Image, and Anchor Box Optimization
Full text linkThe classification and detection of areca nuts are essential for agriculture and food processing to ensure product quality and efficiency. The manual classification of areca nuts is time-consuming and prone to human error. For a more accurate and efficient automated approach, a deep learning-based framework was proposed to address these challenges. This study optimizes the Faster R-CNN by integrating Haar-like features and integral images to enhance object detection. However, dataset limitations, including low image quality, inconsistent lighting, cluttered backgrounds, and annotation inaccuracies, affect the model performance. In addition, the small dataset size and class imbalance hindered generalization. The Faster R-CNN model was trained with and without Haar-like Features and Integral Image enhancement. Performance was evaluated based on training loss, accuracy, precision, recall, F1-score, and mean average precision (mAP). The effects of the dataset limitations on detection performance were also analyzed. The optimized model achieved better stability, with a final training loss of 0.2201, compared to 0.1101 in the baseline model. Accuracy improved from 62.60% to 73.60%, precision from 0.6161 to 0.7261, recall from 0.3094 to 0.4194, F1-score from 0.2307 to 0.3407, and mAP from 0.1168 to 0.2268. Despite these improvements, dataset constraints remain a limiting factor. While the integration of Haar-like features and integral images into faster R-CNN contributes to detection accuracy, the study also reveals that high-resolution images, precise annotations, and dataset scale significantly amplify model performance
Minangkabau Language Stemming: A New Approach with Modified Enhanced Confix Stripping
Full text linkStemming is an essential procedure in natural language processing (NLP), which involves reducing words to their root forms by eliminating affixes, including prefixes, infixes, and suffixes. The employed method assesses the efficacy of stemming, which differs according to language. Complex affixation patterns in Indonesian and regional languages such as Minangkabau pose considerable difficulties for traditional algorithms. This research adopts the enhanced fixed-stripping method to tackle these issues by integrating linguistic characteristics unique to Minangkabau. This study has three phases: data acquisition, pseudocode development, and algorithm execution. Testing revealed an average accuracy of 77.8%, indicating the algorithm's proficiency in managing Minangkabau’s intricate morphology. Nevertheless, constraints persist, particularly with irregular affixation patterns. Possible improvements could include adding more datasets, improving the rules for handling affixes, and using machine learning to make the system more flexible and accurate. This study emphasizes the significance of customized solutions for regional languages and provides insights into the advancement of NLP in various linguistic environments. The findings underscore the progress made in processing Minangkabau text while also emphasizing the need for further research to address current issues
A Multi-Objective Particle Swarm Optimization Approach for Optimizing K-Means Clustering Centroids
Full text linkThe K-Means algorithm is a popular unsupervised learning method used for data clustering. However, its performance heavily depends on centroid initialization and the distribution shape of the data, making it less effective for datasets with complex or non-linear cluster structures. This study evaluates the performance of the standard K-Means algorithm and proposes a Multiobjective Particle Swarm Optimization K-Means (MOPSO+K-Means) approach to improve clustering accuracy. The evaluation was conducted on five benchmark datasets: Atom, Chainlink, EngyTime, Target, and TwoDiamonds. Experimental results show that K-Means is effective only on datasets with clearly separated clusters, such as EngyTime and TwoDiamonds, achieving accuracies of 95.6% and 100%, respectively. In contrast, MOPSO+K-Means achieved a substantial accuracy improvement on the complex Target dataset, increasing from 0.26% to 59.2%. The TwoDiamonds dataset achieved the most desirable trade-off: it had the lowest SSW (1323.32), relatively high SSB (2863.34), and lowest standard deviation values, indicating compact clusters, good separation, and high consistency across runs. These findings highlight the potential of swarm-based optimization to achieve consistent and accurate clustering results on datasets with varying structural complexity